A Universe of Balance….Everything will be fine

It doesn’t matter how much you want to blame your luck, universe has everything for you. Now I want you to think about  consequences, in form of punishment and rewards and share them on the comments. Did that reward or punishment actually helped you?

There are instances where they may had worked or might had not. Think about the last time when you were lauded, did it improved your performance next time. Or did you improve when you were mortified by your failure.

Punishment And Reward

An Israeli fighter pilot when praised for his combat maneuver failed in the test the next time. And when he was reprimanded for his failed attempt, his scores actually improved next time he performed his maneuvers. Punishment have a drastic psychological effect on us. We suffer and realized that we must never to repeat a mistake. At the same time, it may also tend to demoralize and we question our inner talents. In the same way, rewards motivates us in some cases and in some other, it instills a fear for meeting the expectations, which we may not. So I think it works based on type of person we are or the thing that we do.

Regression Toward Mean

Louis is a gambler. He had placed 5\$ on a coin that it will land on head when flipped. But when flipped, it was a tail and he lost. He played again hoping that next time it will be a head based on a 50% probability. He lost again and played again for the third and the fourth time. He lost everytime he placed his bet on heads. He thought that mathematics have gone bonkers.

Here is a sample space for this event: E={T,T,T,T}

Now what is the probability that next event will be head. Is it less than 50%, more than 50% or exactly 50% as was earlier. I have taken more than 1 case so that we can analyse it in terms of ‘change’. Louis thinks that the probability is less than 50% because there are no heads in the sample space. To his surprise, the probability is still 50% or more than 50%. This is because the regression towards the mean will bring the shift in overall probability of sample space to stabilize at 50%. This means as more and more tails turn up in the sample space, the probability of next head will increase. It may be possible that next 4 events will be heads so that overall probability will be 50%. Now problem lies that total number of events are infinite. We can go on and on recording the frequency distribution of heads and tails and sample space will continue to grow! The solution of the problem lies in the problem itself. If the mean is 50%, then the practical outcome can vary but it will get closer and closer towards the mean. It means, the greater the number of events, more accurately can we predict the mean.

In a pack of card, if our first draw is a queen, then the next card drawn will more likely be a card less than a queen.

Rewards and punishments play a little role in performance in the long run. In some case incentives work and in some they do not. So performance is mostly skill based. Next time you have a bad day, cheer up, regression towards mean will turn everything right.